Characterization And Transcriptional Regulation Of CIS Elemets For MRNA Polyadenylation

mRNA 3' polyadenylation is an important step in eukaryotic pre-mRNA maturation process,involving 3' cleavage and polyadenylation of pre-mRNA.Cleavage and polyadenylation site(pA site)is defined by the core and auxiliary cis elements present in the pre-mRNA.These cis elements are also known as the RNA polyadenylation signal(PAS).The core cis element AAUAAA is the most conserved polyadenylation signal located10-30 bp upstream of the cleavage/polyadenylation(pA)site.In mammals,more than 50%pA sites are defined by the canonical AAUAAA hexamer signal,more than 30% pA sites use1-or 2-base variants of the AAUAAA hexamer signal,and the remaining 10-20% pA sites have no recognizable AAUAAA-like hexamer signal.Moreover,over 50% mammalian genes possess two or more pA sites and udergo alternative polyadenylation(APA),selectively producing one of two or more different mature mRNAs in different tissues,different developmental stages,or in response to different environmental stimuli.However,little is known about the cis elements that define the 10-20% pA sites with no recognizable AAUAAA-like hexamer.Many more auxiliary cis elements for the three types of pA sites are yet to be identified.How selective usage of two or more APA sites is regulated by the“strength” of the core and auxiliary cis-acting elements defining each APA site are yet to be elucidated.In this study,I used bioinformatics,flow cytometry,Tet-on system and dual luciferase reporter system technology to address the above questions.First,I optimized in silico identification of core and auxiliary cis elemnets and developed a bicistronic reporter system for characterizing pA site-defining cis elements and studying regulation of alternative polyadenylation.Second,I used the biscistronic reporter system to investigate the roles of RNA secondary structures in defining pA sites and determining the polyadenylation strength of pA sites.Finally,I also studied the regulatory effects of transcription activity(promoter activity)on the polyadenylation efficiency of pA sites.The main results are summarized as follows:(1)I retrieved all the publically available pig EST(expressed sequence tag)data andtranscriptome data as of Sepemtermber 1,2017,and developed an optimized bioinformatics procedure to identify all the pA sites and analyze distribution of nucleotides around each pA site from the retrieved 1,676,405 ESTs and 3,504,376,496 raw RNA-seq reads.A total of62,409 pA sits were identified: 50.96% of them with an AAUAAA hxamer,38.42% with a 1or 2-bp variant of AAUAAA hxamer,and 10.64% without an AAUAAA-like hexamer.And76.69% of pig genes have two or more than two pA sites.(2)I developed a new bicistronic reporter system composed of two fluorescence vectors and two luciferase vectors that have a common structure frame of 1 promoter + 1st ORF + 2restriction enzyme sites + 1 IRES element + 2nd ORF + 1 restriction enzyme site + 1 pA site(SV40 pA)and another restriction enzyme site.This common structure frame allows insertion of one candidate pA site or cis-acting element downstream of the 1st ORF [humanized renilla luciferase(hRluc)or DsRed] and replacement of the SV40 pA with another candidate pA site or cis-acting element downstream of the 2nd ORF [humanized firefly(hluc)or EGFP]simultaneously.As expected,transfection of the empty luciferase or fluorescence protin vector into mammalian cells produced a bicistronic mRNA with a DsRed/EGFP(or hRluc/hluc)ratio of 1 at both the mRNA and protein levels.By constast,inserting a wildtype synthetic pA(SPA)site or mutant SPA site with a 1-5 base variants of the canonical AAUAAA hexamer element or no hexamer immediately downstream of the 1st ORF yielded a mRNA and protein ratio of DsRed/EGFP(or hRluc/hluc)from 0(wiltytype SPA)to 1(mutant SPA with no hexamer or a hexamer of ?3-base substitutions).Inserting the proximal(pA1),distal(pA2),or both pA sites(200 bp up-and downstream of each cleavage site)of human CD47 into the wildtype fluorescence or wildtype luciferase vector showed that the CD47 proximal pA site was a much stronger pA site than the CD47 distal pA site.Taken together,these results demonstrate that this bicistronic reporter system can be utilized not only to quantify the strength of a single candidate pA site or cis element,but also to accurately measure the relative usage of two APA sites at both the mRNA(qRT-PCR)and protein levels.(3)I conducted a genome-wide in silico analysis of the distribution of RNA secondary structures surrounding the pA sites of human genes.This bioinformatic analysis showed that pA sites with no AAUAAA-like hexamer tend to have the most stable secondary structure motifs,followed by pA sites with an AAUAAA-like hexamer,and those with an AAUAAA hexamer.Luciferase analysis showed that inserting a stem structure known as MS2 upstream of the cleavage site of the SPA site enhanced polyadenylation efficiency,whereas inserting MS2 downstream of the cleavage site of the SPA site reduced polyadenylation efficiency.Point mutation experiments revealed that the up-and downstream secondary structures enhanced the polyadenylation efficiency of Edf1 and Mrpl55 pA sits,reduced that of Actr1 bpA site,and had no significant impacts on those of Hotair and Emg1 pA sites.Taken together,these data suggest that many of the RNA secondary structures located 100 bp up-or downstream the cleave site act as auxiliary cis-acting elements.(4)I conducted both promoter deletion and Tet-on experiments to study transcriptional regulation of polyadenylatiob efficiency of the wildtype and mutant SPA.Promoter deletion expersimets showed that lower transcriptional activity resulted in lower polyadenylation efficiency for both strong and weak pA sites.By contranst,greater transcriptional activity induced by doxycycline increased the polyadenylation efficiency of both strong and weak pA sites.The magnitude of the effects of promoter deletion and doxycycline induction on polyadenylation was positively correlated with the strength of the tested pA sites.In summary,I have improved bioinformatics procedures to identify pA sites and the related core cis elements from pig and human transcriptome data and studid the distribution of nucelotides and RNA secondary structures sourrounding pA site.The pig polyadenylation signal usage patterns revealed here have higher accurancy and coverage than the ones previously reported.The new bicistronic reporter system developed in this study not only can quantify the strength of a single candidate pA site or cis element,but also can accurately measure the relative usage of two APA sites at both the mRNA(qRT-PCR)and protein levels.There are less RNA secondary structures up-and downstream of the cleavage site.But if they exist,they usually serve as auxicllary cis elements to enhance or reduce the polyadenylatin efficiency of the pA site.Another important finding of this study is transcription activity positively regulate polyadenylation efficiency.